mlwp-tools
Tools for machine-learning weather prediction
mlwp-tools is a collection of Python packages for working with machine-learning weather prediction (MLWP) data: defining dataset conventions, loading data into them, aligning datasets in space and time, and verifying forecasts.
Packages
| Package | Description |
|---|---|
| mlwp-data-specs | Dataset specifications for MLWP applications for time, coordinate and uncertainty traits based on xr.Datasets and CF Zarr |
| mlwp-data-loaders | General-purpose data-loaders to create mlwp-data-spec conforming xr.Datasets |
| mxalign | xarray-based space/time/uncertainty-alignment for meteorological datasets |
| mlwp-harp-writer | Write mxalign-aligned, mlwp-data-specs conforming datasets to HARP parquet datasets |
| mlwp-scorecard | Easy weather model scorecard creation in Python |
See the roadmap for where the project is heading.